Unit Economics of Private Label Pet Bags

Pet carrier production desk · Updated 2026-10-07 · 14 min read

Custom pet bag orders are quoted by unit and by filled container, because carton cubage decides the landed cost more often than the fabric does.

The Landed Cost Build-Up, Line by Line

A landed cost is a stack, and the only way to manage it is to see every layer. Brands that track only the ex-works price routinely mis-price a range by 25 percent or more, because the layers above the factory gate are where the money actually goes.

The first layer is material. On a mid-size soft carrier: outer shell 1.1 metres of 600-denier polyester at roughly USD 2.30, lining 0.8 metres at USD 0.90, a floor board and stiffeners USD 1.10, webbing and straps USD 1.40, hardware — zips, buckles, rings, feet — USD 2.20, mesh and trim USD 0.60, thread and consumables USD 0.35. That is USD 8.85 of material before waste, and waste on a cut-and-sew soft product runs 8 to 12 percent, so call it USD 9.70.

The second layer is direct labour: cutting, sewing, assembly, trimming, packing. On a bag with a moderate stitch count this is roughly 22 minutes of operator time across the line, and at a fully loaded shop rate it comes to about USD 5.10. The third is factory overhead and margin, typically 28 to 35 percent on top of material and labour, which puts the ex-works price at roughly USD 19.60 at a 500-piece order.

Then come the layers nobody puts on the price list. Export cartons and packing, USD 0.55 per unit. Inland haulage to port and export clearance, USD 0.30. Sea freight, which for this product is driven by volume rather than weight: a flat-packed soft carrier occupies about 0.021 cubic metres, so at a freight rate of USD 90 per cubic metre that is USD 1.89 per unit, or USD 0.94 if the range is specified to fold flat. Marine insurance, 0.3 percent of invoice value. Duty, which varies by classification and destination: 6.5 percent on the customs value into the EU for most travel-bag classifications, 17.6 percent into the US for the same article under the common polyester travel-good headings.

Add customs clearance, port charges and inland delivery to a domestic warehouse — call it USD 1.10 — and the landed cost at 500 units is roughly USD 26.40 against an ex-works price of USD 19.60. That 35 percent uplift is the number that breaks pricing models built on the factory quote.

Fixed Costs That Behave Strangely

Tooling, sampling and testing are fixed in the sense that they do not vary with units, and they behave strangely in the sense that they are not one-off. Understanding both properties is worth more than a dollar of unit price negotiation.

Tooling on a pet carrier programme is modest compared with a moulded product but not trivial. A set of cutting dies for a medium-size bag runs USD 180 to 320. A custom hardware die for a branded buckle or zip puller is USD 400 to 900. An embroidery digitizing fee for a logo is USD 35 to 90. A printed label or hang tag plate is USD 60 to 150 per colour. A washing or compliance test is USD 250 to 900 depending on scope. A realistic first programme carries USD 900 to 1,800 of these, and at 500 units that is USD 1.80 to 3.60 per unit — a material number on a USD 20 product.

The amortisation curve is brutal at low volume and flat above about 3,000 units. The same USD 1,400 of tooling is USD 2.80 per unit at 500, USD 0.70 at 2,000 and USD 0.28 at 5,000. This is the entire commercial argument for committing to a larger first order, and it is why a brand that orders 500, sells out and reorders 500 pays the tooling twice while a brand that orders 2,000 once pays it once.

The second property — that these costs are not truly one-off — catches people. A cutting die wears and is replaced every 8,000 to 12,000 cuts. A test report expires or has to be reissued when a material changes. A hardware die is durable but the hardware supplier has a minimum order that may be larger than the bag order, so the first programme often carries surplus hardware that has to be stored and re-used, which is working capital sitting on a shelf.

MOQ as a Cash Decision Rather Than a Rule

A 500-piece minimum is usually presented as a constraint, and it is better understood as a cash-cycle instrument. The question is not whether you can sell 500 bags; it is how long 500 bags sit in a warehouse, because sitting inventory is the real cost of a minimum order.

Run it as a simple calculation. At a landed cost of USD 26.40 and a retail price of USD 89, an order of 500 ties up USD 13,200 of cash. If that stock turns in 90 days, the carrying cost at a 12 percent annual cost of capital is about USD 390, or USD 0.78 per unit, plus warehousing. If it turns in 270 days, the carrying cost triples and the per-unit penalty is USD 2.34, which is larger than the unit-price saving you would get from doubling the order.

That comparison produces a counter-intuitive and correct conclusion: a larger order is only better if the sell-through rate supports it. Ordering 2,000 units to save USD 1.20 per unit and then holding the stock for three quarters is a net loss of roughly USD 1.14 per unit against ordering 500 three times. The arithmetic favours the small repeat order unless the demand signal is strong.

Note what the table shows: moving from 500 to 5,000 units saves USD 5.57 per unit, and USD 2.52 of that is tooling. Only USD 3.05 is genuine manufacturing scale. That is the honest picture of volume leverage in this category, and it is much smaller than most first-time buyers assume.

Freight Density: The Cost Everyone Underestimates

A pet bag is mostly air. That single fact moves more margin than any negotiation on the factory floor, and it is routinely left out of the pricing conversation until the first freight invoice arrives.

Work the numbers on a mid-size soft carrier measuring roughly 46 by 28 by 30 centimetres. Boxed for shipment that is 0.039 cubic metres per unit at 2.1 kilograms — a density of 54 kilograms per cubic metre, which is very high for a soft good and keeps freight cheap. But specified to ship flat-packed or nested, the same product occupies 0.021 cubic metres, and a structured carrier that cannot be compressed at all occupies 0.052. Across one order of 500 units, the difference between the best and worst case is 15.5 cubic metres, which at USD 90 per cubic metre is USD 1,395, or USD 2.79 per unit — more than the entire tooling amortisation at that volume.

This is why a design decision taken in phase one has a P&L consequence in phase five. A removable floor board, a fold-flat gusset and a nesting pattern are freight measures, and they are worth specifying even when they add 30 cents of construction cost. Conversely, a rigid shell or a fixed internal frame should be chosen with the freight cost already in the model, because on a bulky product it can add 12 to 18 percent to landed cost. Our note on export carton planning covers how carton size is optimised against both the product and the container.

Air freight deserves a specific warning. It is charged on the greater of actual and volumetric weight, using a divisor of 6,000, so a 2.1 kilogram bag occupying 0.039 cubic metres bills as 6.5 kilograms — three times its real weight. At an air rate of USD 4.20 per kilogram that is USD 27.30 per unit of freight on a product with an ex-works cost of USD 22. Air freight is not a logistics option for this category except as a partial, emergency quantity.

Contribution Margin by Channel

The same bag produces four different margins depending on where it is sold, and the differences are large enough that channel choice is a bigger pricing decision than cost negotiation. Contribution here means retail price less landed cost less the channel-specific cost of the transaction.

Direct to consumer through your own site: retail USD 89, landed USD 26.40, payment processing 2.9 percent plus a fixed fee of about USD 0.90, pick-pack-ship USD 6.20, returns provision at a 9 percent return rate and a USD 11 cost per return event, and customer acquisition which at a mature pet accessories brand runs USD 14 to 26 per order. Contribution before acquisition is roughly USD 47; after acquisition at USD 18 it is about USD 29, or 33 percent of retail.

Through a marketplaces channel: the same bag at USD 79, a referral and fulfilment fee of 15 to 17 percent plus a fulfilment charge of USD 8.40 for a bulky item, and a higher return rate at 13 percent. Contribution lands around USD 24, or 30 percent, and the storage fee on an oversize item through a peak quarter can take another USD 2.

Through a wholesale or retail account: the bag sells to the retailer at USD 44.50, which is a 50 percent retailer margin on an USD 89 retail. Landed cost USD 26.40, no pick-pack, no acquisition, but a co-op and markdown allowance of 3 to 5 percent and payment terms of 60 to 90 days. Contribution is roughly USD 15.30, or 34 percent of the wholesale price — which is a higher percentage and a much lower absolute number, and requires no marketing spend at all.

The strategic read is that wholesale looks worst per unit and is often best per programme, because it moves volume without acquisition cost and it amortises tooling faster. The right answer for most launches is a wholesale anchor order plus a DTC channel for margin and content, sized so that the anchor order covers the fixed costs.

Break-Even Volume and the Second Order Problem

Break-even on a pet bag programme is not a sales number, it is a contribution number, and it should be calculated against cash rather than against accounting profit. The distinction matters because the fixed costs are paid in cash months before the revenue arrives.

Take the indicative programme: USD 1,400 of tooling and testing paid at the start, USD 3,300 deposit paid at order placement under T/T 30/70, and the balance of USD 7,700 paid before shipment. Cash out before revenue is USD 12,400. At a DTC contribution of USD 29 per unit after acquisition, break-even is 428 units — which means the entire first order of 500 is consumed reaching break-even, and no profit exists in season one.

That is normal and it is fine, provided it is known. The mistake is treating the first order as a profit event rather than as a market entry. The profitable unit is the second one, because it carries no tooling, no sampling and a shorter specification cycle, and the fixed cost on a reorder is typically less than USD 150 of document refresh. Contribution on a reorder rises by USD 2.50 to 3.00 per unit purely from the absence of fixed cost.

So the real planning question is not "what is our margin on the first order" but "how fast can we get to the second". A programme that reorders at 90 days is materially more profitable over two years than one that orders double the quantity and reorders at 12 months, even though the second looks better on a unit-cost spreadsheet. The reorder mechanics — what is retained, what is re-tested, what changes — are set out in the reorder process. Where the first order needs to be split across sizes rather than colourways, splitting the MOQ changes the amortisation again, because each size carries its own cutting die.

Price Ladder Architecture Across a Range

A range is priced as a ladder, and the ladder has to be built so that the customer steps up rather than steps out. Three errors recur and all three are avoidable at specification stage.

The first is a linear ladder. Prices of 69, 89, 109 and 129 look orderly and they produce a poor mix, because the gap between adjacent steps is small relative to the perceived difference, so buyers cluster at the bottom and the range under-delivers. A geometric ladder with widening gaps — 69, 94, 129, 179 — works better, because the step-up cost grows with the step-up value and the top of the range stops cannibalising the middle.

The second is a ladder where the price steps are not matched by visible value steps. If the USD 129 bag differs from the USD 94 bag only in size, customers will buy the USD 94 one, because size differences are hard to perceive online and easy to rationalise away. Each rung needs a visible, photographable difference: a different material, a different hardware finish, a structural feature, a bundle. Size alone is the weakest possible rung.

The third is forgetting that the ladder has to work in wholesale too. A retailer taking three facings will take the bottom, the middle and one step above, and if the wholesale margin structure is flat across the ladder, the retailer has no incentive to promote the expensive one. A ladder where the percentage margin rises by 2 to 3 points per rung gives the retailer a reason to sell up, and it costs the brand very little because the absolute margin still rises.

The Five Variables That Move Margin Most

If you have one hour to improve the economics of a programme, spend it on these five in this order. Everything else is second-order.

The first is shipped cube, and it is worth between USD 1.50 and USD 3.00 per unit. The second is the number of distinct colourways and sizes, because each variant is a setup and a cutting run, and collapsing four colourways to two with a third introduced at reorder typically saves USD 0.90 to USD 1.60 per unit. The third is hardware: a custom branded buckle at USD 1.85 against a stock buckle at USD 0.42 is USD 1.43 per unit, and it is only worth it if the hardware is the reason someone buys.

The fourth is the return rate, and it is the most under-modelled number in the category. A 9 percent return rate at a USD 11 processing cost and a 60 percent resale recovery costs about USD 1.05 per unit sold; a 15 percent rate costs USD 2.75. Fit specification is therefore a margin lever, not a design nicety — and the single cheapest intervention is a size chart built on interior dimensions, which we cover in sizing development.

The fifth is the reorder interval, for the reason set out above: a reorder carries almost no fixed cost, so every week saved on the reorder cycle is margin. Practically, that means retaining the approved sample, retaining the tooling, and not changing the material between orders unless there is a reason that is worth re-testing for.

How the compliance line sits in the model

Compliance is usually treated as overhead, and it is better treated as insurance with a known premium. Testing a mid-size programme into two markets costs USD 600 to 1,800, which is USD 1.20 to 3.60 per unit at 500 pieces, and it is the difference between a range that can be listed and one that cannot. The cost of getting it wrong is not the test fee; it is a withdrawn listing, which converts the entire first order into dead stock at the worst possible moment in the cash cycle.

Where the range is sold into the United States, the chemical and mechanical evidence is framed against US Consumer Product Safety Commission expectations; where it is sold into the European Union, the same work is framed against the EU REACH framework. The two are not interchangeable and a programme covering both markets needs both, which is why the market list belongs in the cost model rather than in a later conversation. The document set itself is described in the compliance pack.